Podcasts are unique forms of unimodal modality because they include features like conversation, description, and sound effects to encourage audio engagement. Research shows that learners benefit from learning in two modalities (audio + visual) when information is complementary, not redundant. However, these previous studies used audio narration of text as auditory stimuli which differs from podcast formats. Do children learn from podcasts, and does providing supporting visual information affect learning? Children listened (or listened and viewed related images) to an 11-minute science podcast and answered recall and transfer questions. There was no effect of modality on children's learning, and children in both conditions performed above chance on transfer questions. Using a semantic textual similarity analysis, we show that children in the audiovisual condition do not incorporate visual information in their description of concepts. These results highlight the uniqueness of podcasts as a unimodal context that could benefit higher-order concept learning.